Permian–Triassic red-stained albitized profiles in the granitic basement of NE Spain: evidence for deep alteration related to the Triassic palaeosurface

Extensive areas of the Variscan granitic basement in NE Spain display profiles of red-stained albitized facies characterized by albitization of Ca-plagioclase, chloritization of biotite and microclinization of orthoclase, along with the alteration of igneous quartz to secondary CL-dark quartz. These profiles have a geopetal structure beneath the Triassic unconformity, with a very intense and pervasive alteration in the upper part that progressively decreases with depth to 150–200 m where the alteration is restricted to the walls of fractures. The red albitized facies contains secondary maghemite and hematite that indicate oxidizing conditions. Dating of microclinized orthoclase and secondary monazite that have formed in the red-stained albitized facies yielded K–Ar and U–Th–Pbtotal ages of 240 and 250 Ma, respectively, suggesting that the alteration developed during the Permian–Triassic period. The geopetal disposition of the red albitized profile with respect to the Triassic unconformity, its large regional extent, and the fracture-controlled alteration in the lower part of the profile indicate groundwater interaction. The δ18O values of albitized plagioclase (+ 11‰), microclinized orthoclase (+ 13‰), and secondary CL-dark quartz (+ 12‰) suggest that the alteration temperature was about 55 °C. This “low” temperature suggests that the alteration occurred during interaction of the granitic rocks with Na-rich fluids below a surficial weathering mantle on the Permian–Triassic palaeosurface. The latter is possibly related to Triassic evaporitic environments in long-lasting, stable landscapes in which Na-rich solutions infiltrated deep regional groundwaters.

orthoclase, along with the alteration of igneous quartz to secondary CL-dark quartz. These profiles have a 23 geopetal structure beneath the Triassic unconformity, with a very intense and pervasive alteration in the 24 upper part that progressively decreases with depth to 150-200 m where the alteration is restricted to the 25 walls of fractures. The red albitized facies contains secondary maghemite and hematite that indicate 26 oxidizing conditions. Dating of microclinized orthoclase and secondary monazite that have formed in the 27 red-stained albitized facies yielded K-Ar and U-Th-Pb total ages of 240 and 250 Ma, respectively, 28 suggesting that the alteration developed during the Permian-Triassic period. The geopetal disposition of 29 the red albitized profile with respect to the Triassic unconformity, its large regional extent, and the 30 fracture-controlled alteration in the lower part of the profile indicate groundwater interaction. The δ 18 O 31 Introduction alteration event, 3) evaluate the origin and temperature of the albitizing fluid, and 4) gain an 73 understanding of subsurface processes that operated beneath the Triassic palaeosurface. 74 Geological setting and study sites 75 The granitic massifs of the Guilleries (Catalan Coastal Ranges) and Roc de Frausa (Eastern Pyrenees) 76 form part of the Variscan crystalline basement of NE Spain (Fig. 1a, b)  Methods and materials 122 The analytical data for this study are included in Supplementary Information S2 and S3 and are 123 available also in the institutional repository [dataset] Fàbrega et al. (2018). 124 Sampling and mapping 125 Sampling, mapping and construction of cross-sections of the red albitized granitoids were systematically 126 carried out along the Triassic unconformity of both massifs (Fig. 2). Due to a dense forest cover in most 127 of the study areas, the work was mostly carried out along linear features including roads, trails and creeks. 128 A total of about 280 and 100 samples were collected in the Guilleries and Roc de Frausa Massifs, 129 respectively, including pervasively albitized rocks, albitization restricted to fractures and unaltered rocks 130 (Table 1). 131 Table 1. Sampling sites, their numbering in Figure 1, nature of the host rocks and their alteration features. 132 massif sampling sites N. in (Fig. 1 Petrological and geochemical analyses 134 About 80 samples were prepared as thin sections and epoxy-embedded mounts for petrographical 135 analyses. Samples were observed by optical microscopy and optical cathodoluminescence (CL) using a 136 Technosyn Cold Cathodoluminescence 8200-MKII-CL operating at 15-18 kV gun potential and 150 -137 350 μA beam current at the Facultat de Ciències de la Terra of the Universitat de Barcelona (Spain). Frausa and on similar red-stained albitized rocks. In total about 1400 spots were analysed. 153 The degree of microclinization of 73 K-feldspar phenocrysts was quantified by X-ray diffraction (XRD) 154 at the Center of Geosciences, Mines ParisTech, Fontainebleau (France). Identification of the orthoclase, 155 microcline and albite was carried out by peak deconvolution using the Fityk software (Wojdyr, 2010) 156 adjusting the shape of the deconvoluted peaks by the Pearson VII function (Hall et al., 1977). The water and acetone. The K content was measured by XRF on 50 mg aliquots using a specific regression for 163 measuring K in K-Ar samples (Solé and Enrique, 2001). Analytical precision was > 2%. Duplicate 164 samples weighing between 1 and 2 mg were mounted on holes of a copper plate. This plate was placed on 165 an ultra-high vacuum chamber that was degassed at ~150°C for twelve hours before analysis to reduce 166 atmospheric contamination. Argon was extracted by complete sample fusion using a 50 W CO 2 laser 167 defocused to 1 -3 mm diameter. The evolved gasses were mixed with a known amount of 38 Ar spike and 168 purified with a cold finger immersed in liquid nitrogen and two SAES getters in a stainless-steel 169 extraction line. Measurements were done in static mode with an MM1200B noble gas mass spectrometer 170 using electromagnetic peak switching controlled by a Hall probe. Analytical precision on 40 Ar and 38 Ar 171 peak heights was better than 0.2%, and better than 0.5% for 36 Ar. The data were calibrated with internal 172 standards and the international reference materials LP-6 biotite, B4M muscovite and HD-B1 biotite. All 173 ages were calculated using the constants recommended by Steiger and Jäger (1977). A detailed 174 description of the procedure and calculations is given by Solé (2009 In the unaltered igneous rocks, primary monazite occurred as euhedral (20 -50 µm) or rounded (usually 193 150 -200 µm) grains that did not show internal BSE-zoning. In the red-stained albitized granitoids, 194 monazite is exclusively preserved in a mm-scale rock volume attached to the reaction front. In this 195 volume, monazite is pseudomorphosed by secondary monazite and apatite, and displays dissolution-196 reprecipitation textures including corroded grain boundaries and widespread µm-size porosity (Putnis, 197 2002;Harlov et al., 2011). 198 The analyses were performed on three representative samples of unaltered rock and two representative 199 samples plus 2 subsamples of red albitized rock. A total of 530 spots were analysed. Afterwards, the 200 EPMA raw U-Th-Pb total results presenting a sum of oxides less than 95% and/or a relative standard 201 deviation of U and/or Th and/or Pb larger than 20% were rejected. Finally, 117 and 80 analyses were used 202 to estimate the U-Th-Pb ages for the Guilleries and Roc de Frausa, respectively. The U-Th-Pb total dates of 203 the individual EPMA points were calculated using the MonaziteAge software included in the JEOL JXA-204 8230 electron probe, which uses the approach of Adachi (1991, 1994) and Suzuki et al. 205 (1994). Subsequently, the standard error of each single date was estimated using the Microsoft Excel 206  (Ludwig, 2003). 210 The IMF of the plagioclase, K-feldspar and quartz δ 18 O SIMS analyses were predicted by response 236 surface methodology (RSM) using the three response surface models described in Fàbrega et al. (2017). 237

In-situ δ 18 O SIMS analyses
The predictor (input) variables were the instrumental parameters X and Y stage position, primary beam 238 intensity (PI), chamber pressure (CP) and, electrostatic deflectors LT1DefX and LT1DefY. In addition, 239 the models for feldspars included the compositional inputs An% (plagioclase model) and Or% and BaO% 240 (K-feldspar model). The IMF was considered as the predicted (output) variable or response. This RSM 241 approach permitted to predict a unique IMF value for each SIMS analysis. The unaltered facies in Roc de Frausa is mainly porphyritic granite with K-feldspar phenocrysts (2-3 cm), 265 and this also occurs in the St. Aniol, Hortmoier and Oix areas (Fig. 2c). Primary Ca-plagioclase occurs as 266 zoned euhedral 0.5 -2 mm crystals with polysynthetic twinning, yellow-green luminescence and a 267 composition in the ranges Ab 66-74 and Ab 90-99 for the Ca-rich cores and the Na-rich zones, respectively 268 ( Table 2). The matrix K-feldspar is subhedral to euhedral 0.5 -2 mm crystals with micro-and crypto 269 perthite lamellae, blue luminescence and a composition of Or 74-98 ( Table 2). The K-felspar phenocrysts 270  The reactions that affect the major minerals are albitization of Ca-plagioclase, chloritization of biotite, 312 microclinization of orthoclase, and the alteration of igneous quartz to secondary CL-dark quartz (Fig. 4). 313 Primary Ca-plagioclase is pseudomorphosed by secondary albite. Albitized plagioclase grains ( Fig. 5a, b, 314 c) have a turbid aspect, a lack of luminescence, widespread micron-sized non-connected euhedral porosity 315 ( Fig. 5d) and a very pure chemical composition in the range of Ab 97-99 (Table 2). The replacement front 316 between the primary Ca-plagioclase and the secondary albite is very sharp (Fig. 5b), and the 317 crystallographic framework is preserved from the primary to the secondary phase. These textural features 318 are typical of a fluid-aided dissolution-reprecipitation process along micro-fractures without development 319 of dissolution pores prior to precipitation (Putnis, 2002;Engvik et al., 2008;Putnis, 2009). The albitized 320 plagioclase grains usually contain micron-sized non-luminescent secondary K-feldspar and orange-CL 321 calcite (Fig. 5c), the latter with a composition about Magnesite 1.5 Siderite 0.7 Calcite 97.8 . The primary 322 green-yellow luminescent Na-rich rims of plagioclase, with the composition Ab 90-93 , are preserved during 323 albitization (Fig. 5a). 324 Primary biotite is pseudomorphosed by secondary chlorite. Chloritized biotite is non-luminescent, 325 preserves the biotite sheet structure and usually contains lens-shaped inclusions of secondary K-feldspar, 326 quartz and orange-CL calcite (Fig. 5c) The chloritized biotite grains systematically show a significant increase of yellow luminescent micron-330 sized fluorapatite inclusions (Fig. 5c). In places, neoformed chlorite occurs in sheaf like arrangements of 331 micron-size sheets inside the micro porosity. 332 The microclinization of orthoclase is characterized by the recrystallization of primary orthoclase to 333 secondary microcline. The microclinized grains have a cloudy aspect, loss of luminescence (Fig. 5a), a 334 widespread micron-sized non-connected euhedral porosity, amoeboid-shaped patchy perthite texture and 335 a composition of about Or 96-99 and Or 87-93 in the Guilleries and Roc de Frausa, respectively ( Table 2). The 336 patch perthite exclusively forms during the microclinization process and implies replacement of K-337 feldspar by secondary albite. K-feldspar phenocrysts (2 -3 cm) in the porphyritic granites progressively 338 change from the white color of the primary orthoclase to brown and finally red-pink as the degree of 339 microclinization increases. The most strongly microclinized phenocrysts have a composition (XRD) of 340 about 40 -60% microcline, 20 -30% albite and 10 -40% orthoclase. The albite detected by XRD inside 341 the red-pink phenocrysts includes the patch perthite and plagioclase inclusions that were albitized during 342 microclinization. 343 The net of micro fractures in the red-stained albitized rocks is sealed by secondary CL-dark quartz that 344 postdates the aforementioned reactions. Within quartz grains, this net of micron sized CL-dark quartz has 345 a dendritic texture (Fig. 5e), suggesting that it propagated along sub-micron cracks and dislocations and 346 may possibly record some degree of dissolution-reprecipitation process of the primary quartz. Monazite is strongly altered in the red and pink facies of the upper part of the profile, where the grains are 373 completely pseudomorphosed by apatite and quartz inside a millimetre-scale zone adjacent to the reaction 374 front (Fig. 5f). The secondary monazite shows evidence of dissolution-reprecipitation mechanisms 375 including corroded boundaries, fracturing, and secondary micron-sized porosity (see for example Putnis, 376 2002Putnis, 376 , 2009). The pseudomorphosed monazite grains are usually accompanied by micron sized authigenic 377 titanite, and synchysite (Fig. 5f). 378 The alteration of all silicates and non-silicates, namely albitization of Ca-plagioclase, microclinization of 379 orthoclase, chloritization of biotite, and precipitation of secondary CL-dark quartz, iron oxides and REE-380 minerals, decreases from the top to the bottom of the profile (Fig. 4). In the same context, the succession 381 of mineral alteration adjacent to fractures decreases in intensity towards the cores of the granite blocks. 382 Decimetre-thick alteration zones occur adjacent to the major fractures, with restricted millimetre-thick 383 zones walls adjacent to secondary fractures, and less along micro-cracks. These relationships point clearly 384 to the influence of fluid circulation. As well, albitization was not associated with a volume change, i.e. the 385 texture and fabric of the altered rocks is not significantly different from that of the unaltered rocks, at 386 least at a macroscopic level. 387 The evaluation of element mobility during fluid-rock interaction can be referred to an "immobile" 392 geochemical framework (Ague and van Haren, 1996). This is usually undertaken using one or more 393 ostensibly immobile elements that are ideally concentrated in minerals that do not change during fluid-394 rock interaction and that present a low analytical uncertainty (Beinlich et al., 2010). 395 To estimate the mass changes caused by albitization (Table 3), the concentrations of the elements in the 396 albitized rock samples were recalculated with respect to Zr, considered as an immobile reference. 397 Petrographic observations show that zircon grains in the albitized rocks remain mostly unchanged from 398 those in the unaltered facies. 399

401
Petrographical observations of the red-stained albitized rocks show that albitization of plagioclase, 402 microclinization of K-felspar, and chloritization of biotite are intimately linked to the precipitation of 403 secondary synchysite-(Ce), apatite, epidote, calcite and hematite that probably recycled some of the ions 404 released by the major reactions. Consistent with the albitization of primary plagioclase, the amount of Na 405 in the albitized rock presents a mass increase of about +91%, and a Ca loss of -60% (Fig. 6). By 406 considering the atomic mass of Na and Ca, this means that only about half of the Ca released by 407 replacement of plagioclase by albite has been removed from the system. The remnant Ca was likely 408 recycled into secondary synchysite-(Ce), apatite, epidote, and calcite. The K content presents a significant 409 reduction of -43% (Fig. 6), which is consistent with the systematic biotite chloritization and some K-410 feldspar albitization. A fraction of the K is probably retained by the secondary K-feldspar lenses within 411 chloritized biotite and micron-size patches of perthite in the albitized plagioclase. Mg presents a 412 significant increase of +30% (Fig. 6). The Si content increases by about +17%, which agrees with the 413 overall albitization of plagioclase (Fig. 6). Al released during albitization of plagioclase may have been 414 consumed during chloritization of biotite and the formation of secondary K-feldspar. However, the 415 albitization reflects an overall increase in Al of +7%. The Fe content remains stable (Fig. 6). The Fe 416 released during chloritization of biotite was probably retained to form the secondary Fe-oxides. The 417 amount of Ti decreased about -5% (Fig. 6), most of it probably released during chloritization of biotite 418 and partly retained in secondary titanite. The variation in Mn shows a loss of -2% (Fig. 6). The loss of P 419 was about -19% (Fig. 6), mostly due to the dissolution of monazite in the albitized profile. Some of the P 420 In the leucogranite of the Guilleries Massif, the zoned plagioclase has δ 18 O values ranging from 8.52 to 432 9.51‰ in the Ca-bearing cores and 11.14 to 11.93‰ in the Na-rich rims (Table 4), showing a positive 433 correlation of δ 18 O values with increasing sodic composition (Fig. 7a). Primary K-feldspar grains have 434 δ 18 O values ranging from 10.42 to 11.65‰ (Table 4), without any correlation with the primary K-feldspar 435 composition (Fig. 7b). Primary quartz has a δ 18 O composition between 6.4 and 9.60‰ (Table 4). In the 436 Tagamanent area (see Fig. 2b), the δ 18 O of the K-feldspar phenocrysts is between 10.6 and 10.8‰, close 437 to the mean value of the primary matrix K-feldspar grains in the Guilleries samples. 438

440
In the porphyritic granite of the Roc de Frausa Massif, δ 18 O values for the primary plagioclase range 441 between 7.18 and 9.45 in the Ca-rich cores and 9.88 to 11.94‰ in the Na-rich rims (Table 4), displaying a 442 positive correlation of δ 18 O values with the increase of sodium content (Fig. 7a). The primary K-feldspar 443 grains in the granitic matrix have δ 18 O values ranging from 11.50 to 12.39‰ (Table 4)  The oxygen isotope fractionation between primary plagioclase and K-feldspar (∆ 18 O Ab-Kfs ) were about 452 +0.5‰ and -0.5‰ for the Guilleries and Roc de Frausa, respectively, suggesting that, within error, 453 primary feldspars formed under equilibrium conditions (Fig. 8a). The oxygen isotope fractionation 454 between primary quartz and primary plagioclase (∆ 18 O Qtz-Pl ) are negative, with values about -3 and -4‰ 455 for the Guilleries and Roc de Frausa, respectively (Fig. 8b). In the case of the quartz and primary K- Albitized plagioclase from Guilleries has δ 18 O ranging from 10.31 to 11.60‰ (Table 4), representing an 471 average increase about +2‰ with respect to the δ 18 O of the primary Ca-bearing plagioclase cores 472 (Fig. 7a). At the Roc de Frausa, δ 18 O values for albitized plagioclase range from 11.14 to 12.75 ‰ (Table  473 4), an average increase about +2.5‰ compared to the primary Ca-rich plagioclase cores (Fig. 7a). 474 The microclinized orthoclase of Guilleries has δ 18 O values ranging from 11.76 to 13.81‰ (Table 4), an 475 average increase about +1.5‰ compared with the primary orthoclase (Fig. 7b). At Roc de Frausa, the 476 δ 18 O composition of microclinized orthoclase ranges from 12.31 to 13.96‰ (Table 4), an average 477 increase about +1‰ compared to the primary orthoclase (Fig. 7b). 478 SEM-CL images of secondary quartz in Roc de Frausa samples showed that the SIMS craters presented a 479 mix of primary and secondary CL-dark quartz. The percentage of secondary CL-dark quartz was 480 estimated using CAD software and the δ 18 O values were plotted against it and regressed to estimate the 481 δ 18 O in both primary and CL-dark secondary quartz (Fig. 7c). Using this method, the primary quartz and 482 secondary CL-dark quartz were estimated to have δ 18 O values of about 7.5‰ and 12‰, respectively 483 (Table 4) Hortmoier (see Fig. 2c). In unaltered rocks, the white primary orthoclase yielded a K-Ar age of 502 283 ± 10 Ma (2σ), which is consistent with the closure of the K-Ar system during magmatic cooling. Solé In the red albitized rocks, brown K-feldspar phenocrysts from albitized fracture walls (see Fig. 3e), 506 represent a degree of microclinization of 30-40% and yielded K-Ar ages of 244.5 ± 6, 237.6 ± 6, and 507 230.7 ± 6 Ma (2σ). The textural similarities and the overlapping of the three ages within the 2σ range 508 support a probable K-Ar resetting event during the Early Triassic. Given that these microclinized brown 509 K-feldspar phenocrysts occur near the reaction front (see Fig. 3e) and are accompanied by partially 510 albitized plagioclase and chloritized biotite, these Early Triassic dates are considered to constrain the age 511 of the alteration in these red fracture facies in the lower part of the albitized profile. 512 In the same locality, the red albitized granite commonly has braided networks of millimetric calcite-filled 513 fractures that cross-cut and thus post-date the albitized plagioclase, microclinized K-feldspar, chloritized 514 biotite and altered quartz grains. K-feldspar phenocrysts in these samples, characterized by intense pink 515 color, yielded K-Ar ages of 216.2 ± 6, 175.9 ± 5, 174.8 ± 5, and 164.0 ± 5 Ma (2σ). These Late Triassic-516 Early Jurassic ages may record an alteration that registers local resets of the K-Ar system relating to the 517 formation of the calcite-filled fractures. 518 519 Monazite U-Th-Pb total dating 520 EPMA U-Th-Pb total dating was carried out on primary monazite grains in the unaltered rocks and on 521 pseudomorphic secondary monazite associated with the reaction front in the red-stained albitized rocks. 522 In the Guilleries Massif (Fig. 9a) in the relicts of primary monazite situated inside the monazite grains pseudomorphosed by secondary 537 monazite (Fig. 10a). Differently, the analyses on the pseudomophic secondary monazite that formed in 538 the albitized rocks near the reaction front (Fig. 10a, b,  If the water-rock oxygen isotope exchange approaches equilibrium (i.e. forward and backward isotopic 589 exchange occurs between the rock and fluid), the oxygen isotope composition of the fluid after re-590 equilibration with the rock (or a given mineral) can be estimated using a mass-balance equation (Taylor, 591 1977 (1) δ 18 O w f and δ 18 O w i being the initial and final oxygen isotope compositions of fluid, respectively, R and W 593 the percentage of oxygen atoms in the rock (or mineral) and water, respectively, and δ 18 O m i and δ 18 O m f the 594 oxygen isotope composition of the initial and final minerals, respectively. 595 Because the isotopic fractionation at equilibrium between minerals and fluid is also temperature 596 dependent (Zheng and Hoefs, 1993), then the equilibrium temperature can be estimated using the oxygen 597 isotope equilibrium equations of plagioclase (Eq. 2), K-feldspar (Eq. 3) and quartz (Eq. 4) with fluid, 598 being (Zheng, 1993  1999a), the Alps (Barth, 2000) and the Fennoscandian and Ukrainian shields (Stober and Bucher, 1999b), 615 the North American Canadian shield Fritz, 1982, 1987;Frape et al., 1984a;Stober and Bucher, 616 1999b), the African Ahaggar Massif and Dodoma area (Nkotagu, 1996;Saighi et al., 2001), and the 617 Indian Ranchi area (Saha et al., 2001), suggesting that negative δ 18 O values are intrinsic of deep 618 groundwater (Kloppmann et al., 2002). Assuming an analogy between the current and the Permian-619 Triassic granitic basement, negative δ 18 O values between -8 and -12‰ can be reasonably assumed for the 620 Permian-Triassic palaeogroundwater in the granitoids of the Guilleries and Roc de Frausa Massifs. 621 The rock to water (R/W) oxygen mass ratio is also a key factor for an estimation of the isotopic evolution 622 of the fluid. Reactions arising directly at fracture walls, where solution may be renewed by circulation, 623 may correspond to relative low R/W ratios, whereas those occurring away from the fractures, at the 624 reaction front within primary crystals in an almost closed system, most likely correspond to higher R/W 625 ratios. To determine a minimum and conservative estimate of the δ 18 O evolution of water, calculations 626 were carried out applying R/W values of 0.5 and 1, considered to be high (≥1) water to rock ratios 627 (Taylor, 1977). After calculation of the resulting oxygen isotope composition of fluid for each reaction, the corresponding temperatures were calculated using Eq. (2), (3) and (4) for albitization of plagioclase, microclinization of 654 K-feldspar and the CL-dark quartz formation, respectively. 655 The results of the calculations are summarized in Table 5  can be considered, within its error estimation, as an isothermal process around 55ºC. 673 Regardless of the lack of precision, these values show that the temperatures of the reactions associated 674 with development of red-stained albitized rocks are significantly below temperatures suggested for 675 hydrothermal albitisation (Cathelineau, 1986;Boulvais et al., 2007) and tardi-magmatic alteration (Lee 676 and Parsons, 1997;Fiebig and Hoefs, 2002), and are consistent with shallow near-landsurface conditions. 677  halite moulds in transgressive epicontinental and continental clastic deposits (Courel, 1982;Galán-691 Abellán and Martínez, 2018). It is possible that in continental settings halite may have been derived via 692 aeolian processes from marine-derived saline lakes and saltflats together with red desert dust typical of 693 Triassic red sandstone deposits (Ruffell and Hounslow, 2006). Leaching of the salt accumulated in the 694 landscape would have provided Na-rich solutions depleted in K relative to marine brines (Fig. 11a).

700
A geochemical mass balance for the studied Triassic red-stained albitized facies (Fig. 6) shows that 701 albitization of plagioclase is triggered by an interaction between rock and Na-rich fluids with replacement 702 of Ca by Na. Salt accumulated in the Triassic landscape is the most likely source of Na-rich solutions. For 703 example, dense saline brines could have been a major contributor to the regional groundwater ( This loose regolith material was later eroded, most probably during the period of relief rejuvenation and 737 intra-belt basin erosion during the Early Triassic (Bourquin et al., 2011) corresponding to the widespread 738 unconformity observed in the German Triassic (Trusheim, 1961;Wolburg, 1968;Röhling, 1991). Erosion 739 and burial beneath Triassic deposits was driven by tectonic activity and sea level rise (Fig. 11b). 740

741
A widespread and distinctive red stained and albitized granitic facies beneath the Triassic unconformity in 742 NE Spain was formed in the Late Permian to Early Triassic based on three independent dating methods. 743 This major result indicates that the albitization and associated alteration are related to the Triassic 744 palaeosurface that unconformably overlies the albitized profile and developed in response to pervasive 745 infiltrating fluid flow while the Variscan granites of NE Spain were exposed at the land surface. The significant amounts of secondary apatite formed during this alteration must be taken into account if 764 fission track analysis or (U-Th)/He dating is carried out on these red-stained albitized rocks. 765